Comparing surgical outcomes of anterior capsular release vs circumferential release for persistent capsular stiffness

Purpose To consolidate the existing literature evaluating anterior capsular release and circumferential capsular release in the treatment of adhesive capsulitis (AC) of the shoulder. Methods The electronic databases PUBMED, EMBASE, MEDLINE and CENTRAL (Cochrane Central Register of Controlled Trials) were searched from data inception to October 8, 2020. Data are presented descriptively where appropriate. A meta-analysis was conducted for patient-reported outcomes. Results Overall, there were forty-six articles included. The majority of patients underwent circumferential release compared to anterior release (80.1% vs. 19.9%). Concomitant Manipulation Under Anesthesia (MUA) was employed in 25 studies, with a higher occurrence in the anterior compared to the circumferential release group (70% vs 60%). Both groups experienced significant improvements postoperatively in range of motion (ROM) and patient-reported outcomes. Complication rates were low for both anterior release (0.67%) and 360° release (0.44%). Conclusion Both anterior and circumferential release are effective techniques for treating AC with low complication rates. Future studies should improve documentation of patient demographics, surgical techniques and outcomes to determine an individualized treatment protocol for patients. Level of evidence Level IV, Systematic Review of Level I–IV studies.


Introduction
Adhesive capsulitis (AC) is characterized by progressive loss of active and passive motion of the shoulder, leading to stiffness and pain. 1,2 This condition is known to affect 2-5% of the general population and a higher proportion of patients with diabetes (10-36%). 3,4 In addition to diabetes, risk factors for developing AC include trauma, thyroid disease and female sex. 1 In cases where conservative treatment fails, surgical management is considered. This includes manipulation under anesthesia (MUA) often in combination with arthroscopic capsular release. MUA alone results in improvements in shoulder motion and function between 6 to 9 months from the onset of symptoms but places a patient at risk of injury to soft tissues or fracture, 5 MUA is thus accompanied generally with arthroscopic capsular release to reduce the risk of iatrogenic injury with the literature supporting significant improvement seen in long-term outcomes with respect to clinical outcome scores and range of motion (ROM). 6 Often, the anterior structures of the capsule are released arthroscopically. However, an extended release of the capsule, also known as a circumferential release (360°) can also be used. 7 Employing the circumferential release technique avoids the need for manipulation in the majority of cases, which is routinely carried out during anterior capsular release, and may presents a potential for complication. 8 These include rotator cuff tear, humeral and glenoid fractures and nerve damage. 9,10 However, there is concern that release of the entire capsule could be associated with instability and potential axillary nerve damage during release of the inferior capsule. 8,11 Outcomes of both procedures have been demonstrated to be effective in improving ROM with low revision and complication rates. 12,13 It remains controversial as to whether an extended capsular release results in improved outcomes over isolated anterior release. 10 The aim of this systematic review is to evaluate available literature to determine if there is any benefit to 360°capsular release over anterior capsular release with respect to ROM, functional outcomes as well as risk of complications.

Methods
This systematic review was conducted according to the methods outlined in the Cochrane Handbook for Systematic Reviews and reported following the Preferred Reporting for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. 14,15 Search strategy PUBMED, EMBASE, MEDLINE and CENTRAL were searched on October 8, 2020 for literature on anterior and 360°capsular arthroscopic release for AC. The search terms included "shoulder", "AC", "anterior capsular release", "360°capsular release" and similar phrases (Appendix I). MeSH and EMTREE terms were used in various combinations and supplemented with free text to increase search sensitivity. The search terms were entered onto Google Scholar and ClinicalTrials.gov, to ensure that relevant articles were not excluded. A hand search of the references of included articles was performed to ensure no potentially eligible articles were missed.

Eligibility criteria
Inclusion criteria were: (1) anterior capsular release or 360°c apsular release; (2) AC or frozen shoulder; (3) at least one outcome reported and stratified for population of interest;

Study screening
The study screening was performed in duplicate by two independent reviewers (S.E., S.S.). Discrepancies that occurred at the title and abstract stage were resolved through discussion and consensus. If consensus could not be reached a senior reviewer was consulted when necessary (A.S.).

Quality assessment
Using the Journal of Bone & Joint Surgery (JBJS) classification system for literature in the field of orthopaedics, the two reviewers determined the level of evidence (I to IV) for each study independently and in duplicate. The methodological quality of non-randomized comparative studies was evaluated using the methodological index for nonrandomized studies (MINORS). A score of 0, 1 or 2 is given for each of the 12 items on the MINORS checklist with a score of up to 16 for non-comparative studies and 24 for comparative studies. Methodological quality was categorized a priori as follows: a score of 0-8 or 0-12 was considered poor quality, 9-12 or 13-18 was considered fair quality, and 13-16 or 19-24 was considered excellent quality, for non-comparative and comparative studies, respectively. The Cochrane Risk of Bias tool was used to evaluate the quality of randomized trials. The Cochrane Risk of bias tool evaluates studies in 7 domains (i.e. random sequence generation, allocation concealment, selective reporting, blinding of participants and personnel, blinding of outcome assessment, outcome data, and other biases) as having high, unclear, or low risk of bias.

Data abstraction
Two reviewers (S.E., S.S.) independently abstracted relevant data from included articles and recorded the data onto an Excel Spreadsheet designed a priori.

Study characteristics
A total of 1093 studies were retrieved from the initial search and following title and abstract screening, 126 studies underwent full-text review. Of these studies, 46 studies met inclusion criteria ( Figure 1). Studies were published between 1989 and 2020, most commonly from the USA (15.2%, n = 7), the United Kingdom (13.0%, n = 6), Korea (10.9%, n = 5) and Japan (10.9%, n = 5). Within the 46 studies, a total of 2261 patients were included with a mean age of 54.0 (SD 5.0) and a mean follow-up of 29.3 months (SD 19.0). Within the overall study group, 80.1% (n = 1810) patients underwent circumferential release in 30/46 studies with a mean age of 54.21 (SD 4.4) and mean follow-up of 30.9 months (SD 17.7). In 10/46 studies, there were 19.9% (n = 451) patients who underwent an anterior/selective release with a mean age of 52.5 (SD 6.1) and a mean follow-up of 32.3 months (SD 23.0) ( Table 1).

Study quality
In this review, the 46 studies include two (4.3%) studies with Level I evidence, 16 (34.8%) Level II evidence, 18 (39.1%) Level III evidence and 10 (21.7%) Level IV evidence (Table 1). Agreement between reviewers was substantial at the title and abstract screening stage (κ = 0.79, 95% CI 0.72-0.85) and excellent at the full text screening stage (κ = 0.91, 95% CI 0.84-0.99). The mean MINORS score for non-randomized comparative studies was 17 (SD 2.2) and for non-comparative studies 12.2 (SD 1.8), indicating fair quality. The RCTs had low risk of bias (Appendix 2). There was excellent inter-rater agreement for quality assessment according to the MINORS criteria (ICC = 0.99, 95% CI 0.98 to 1.0).

Complications
The overall complication rate among all patients included in this systematic review was 0.84% (n = 11). Recurrence of AC was reported in nine cases (0.40%) (8 circumferential and 1 anterior capsular release).

Discussion
The most significant finding of this systematic review was that there were no clinically significant differences found in outcome measures at different post-operative time points between anterior and circumferential capsular release patients. Also of significance is the relatively low complication rate in both the anterior and circumferential release groups (0.67% vs 0.44%). Furthermore, significant postoperative improvements were seen in both groups of patients, particularly in the postoperative outcomes of VAS for pain, Constant scores, and ROM in forward flexion, abduction, internal rotation and external rotation. Forward elevation and UCLA scores were additionally found to be significantly improved for patients who underwent circumferential capsular release however these differences are unlikely to be clinically significant given they did not reach the MCID for the UCLA. 60 The pooled analysis of studies directly comparing anterior capsular release with 360°release revealed no significant differences between the two surgical techniques at all postoperative time periods (e.g. 3 months, 6 months, 12 months, final follow-up). Overall, both procedures provide clinically similar outcomes with respect to function and ROM. Rates of recurrence were also low for both anterior and circumferential techniques (0.04% vs 0.35%).
In the current systematic review, the majority of patients underwent the circumferential capsular release (80.1%). The increased prevalence of circumferential release may potentially be related to the concept of posterior capsular release improving glenohumeral internal rotation which is restricted by the posterosuperior capsule. 58,61,62 It has been reported that circumferential may result in faster recovery, however in a 2015 RCT, there was no significant difference in outcomes between the two techniques at final follow-up. 56 Employing the circumferential release technique also may avoid the need for manipulation in the majority of cases. 12 In this systematic review, the rate of MUA in the anterior release group was higher than the circumferential release group. MUA is routinely carried out during capsular release, and presents a number of possible complications. 12 These include increased risk of dislocation, humeral fracture, rotator cuff tear, brachial plexus injury, joint hemorrhage and inadvertent soft-tissue injury. 12,26,44,57 The low complication rates in both the anterior release technique and circumferential release technique are notable. Controlled capsular release as opposed to isolated MUA decreases the risk of injury to the soft tissue structures and bone and likely results in earlier recovery and improved rehabilitation. 12,44 Although the choice for surgical positioning will largely depend on the surgeon's preferences, there are benefits to both the lateral decubitus and beach chair position. While the lateral decubitus is said to offer improved visualization of the shoulder capsule, the beach chair position allows for intraoperative movement of the humeral head with ease." Recent published results encourage a transition from the more commonly used beach chair position to a lateral decubitus approach due to improved visualization, and potentially overall reduced risk of iatrogenic injury and low complication and revision rates. 11,12 The improved visualization, particularly in the inferior capsule and axillary recess, further eliminates the need for manipulation. 11 The findings of this systematic review do not report differences in patient positioning. Future studies should aim to assess outcomes of anterior release and 360°release based on surgical positioning to determine an optimal treatment method.
Several factors contribute to the post-operative success of arthroscopic capsular release including the timing of surgical intervention and etiology. 49 A recent systematic review found that despite a comparable success rate, diabetic patients experience increased recurrent pain post-surgically and poorer ROM and function compared to idiopathic cases. 63 The arthroscopic capsular release procedure, however, was shown to be equally effective in patient groups classified by etiology including postsurgical, idiopathic, post-traumatic, and diabetic. 53,64 The surgical technique selected for each patient should be dictated by the restriction in ROM exhibited by the patient. 59 The superior and middle glenohumeral ligaments, the rotator interval, the coracohumeral ligament extraarticularly and/or the intra-articular portion of the subscapularis should be released to address loss of external rotation while release of the anteroinferior capsule addresses loss of elevation. 59 A posterosuperior release can be performed to address loss of internal rotation. 59 Limitations Some studies had poor documentation of patient demographics and surgical techniques and had to be excluded from the review. Furthermore, the pooling of outcomes was limited due to the lack of available studies for this purpose. Pooling across multiple studies was not possible due to the large volume of studies included, each reporting different outcome measures. There was a lack of detail provided regarding how MUA was performed between the study groups, limiting our ability to determine its effect on postoperative outcomes and complications. There were also several inconsistencies in complications reporting as some studies did not report complications at all, whereas others reported only major complications leading to revision while excluding minor complications. Thus, the overall complication rates found in this systematic review may be subject to inaccuracies and be possibly underestimated/underreported. Although statistically significant differences were found they were unlikely to be clinically significant. The included studies had varying follow-up periods; thus, it was difficult to ascertain the trajectory of improvement postoperatively. The variability between studies in postsurgical treatment approach, including physiotherapy technique and prescribed analgesic, is notable. As these choices may directly influence the final outcome of the surgical intervention, making direct comparisons was not possible. A lack of documentation in outcomes distinguishing between primary AC and secondary AC patients limited our ability to determine the influence of etiology on outcomes. Thus, there was a limited ability to make adequate comparisons between the two techniques.
Additionally limited high quality studies are available to inform treatment decisions.

Future studies
Future studies should aim to improve documentation of patient demographics, etiology of AC, surgical techniques, rehabilitation protocol and outcomes. The effect of time of onset of symptoms, as well as delay in treatment on patient outcomes should be assessed. This will allow clinicians to determine the ideal time to treat patients based on clinical symptoms and patient history. A comparison between patient positioning and surgical techniques will be needed to determine an optimal treatment method. Future studies should specify the stage of AC to assess outcomes for similar patient groups. Furthermore, the class of stiffness must be defined to ensure consistent conditions in comparing procedures. This can be accomplished using an RCT study design with large sample sizes in order to limit inclusion bias. In addition, a guideline for clinicians considering patient history, imaging and surgical findings should be established to determine an appropriate treatment for patients. The development and implementation of a standardized MUA protocol is needed for research purposes with the intent of comparing outcomes of different surgical techniques. Furthermore, as the current literature reports distinct combinations of outcome measures, a set of standardized core outcomes for AC should be instituted. The differences in postsurgical treatment method likely influenced the outcomes and should be taken into account. Hence, RCTs with a clearly defined physiotherapy and pain protocol are needed to rule out bias associated with post-surgical rehabilitation. Complete reporting of outcomes and complications will permit for a more in-depth analysis and comparison between anterior release and circumferential release.

Conclusion
Anterior capsular release and circumferential capsular release both result in significant improvements for patients with AC with respect to ROM as well as functional outcomes with very low complication rates with both procedures. The most significant finding of this systematic review was that there were no clinically significant differences found in outcome measures at different postoperative time points between anterior and circumferential capsular release patients.